Clinical Question

Is transdermal testosterone replacement safe in men with known heart disease or at high risk of heart disease?

Bottom Line

Although the most serious cardiovascular (CV) events were not more common in high-risk men using testosterone gel, there were higher incidence rates of acute kidney injury, atrial fibrillation, nonfatal arrhythmias, and pulmonary embolism. Testosterone supplementation should be used with caution, if at all, in men with a history of these conditions. (Level of Evidence = 1b−)

Synopsis

Data regarding the safety of testosterone supplementation have been mixed, with some studies finding increased CV risk. The U.S. Food and Drug Administration, therefore, mandated that companies that sell these supplements perform a safety study. The researchers identified men, 45 to 80 years of age, with symptoms of hypogonadism and two fasting morning serum testosterone levels of less than 300 ng per dL (10.41 nmol per L). All of the patients had preexisting CV disease (55%) or were at increased risk of CV disease (45%). The patients were randomized to receive testosterone gel titrated to achieve a testosterone level of 350 to 750 ng per dL (12.15 to 26.03 nmol per L) or matching placebo gel with sham titrations. Testosterone was discontinued if the patient's serum level exceeded 750 ng per dL even at the lowest dose, or if hematocrit level exceeded 54%. Groups were balanced at baseline, with a mean age of 63 years, and 80% were White, 69% had diabetes mellitus, 93% had hypertension, and 90% had hyperlipidemia. Analysis was by intention to treat. The full study population included 5,204 patients; 5,198 received at least one dose of testosterone gel and were included in the safety population for the primary analysis. Patients were treated for a mean of approximately 22 months and followed for a mean of 33 months. There was no difference between groups in the primary safety outcome of major adverse cardiac events, which included CV death, nonfatal myocardial infarction, and nonfatal stroke (7.0% vs. 7.3%). There was also no difference for any of the components of the composite outcome, all-cause mortality, need for revascularization, or need for hospitalization. The authors did see an increase in the risk of nonfatal arrhythmia that warranted an intervention (5.2% vs. 3.0%; P = .001; number needed to harm [NNH] = 45 over 21 months), atrial fibrillation (3.5% vs. 2.4%; P = .02; NNH = 91 over 21 months), and acute kidney injury (2.3% vs. 1.5%; P = .04; NNH = 125 over 21 months). Improvement in symptoms was not reported. Pulmonary embolism was also more common with testosterone (0.92% vs. 0.46%). The authors did not report statistical significance for this outcome.